Looking at 'Primitive Skills' - Iron Smelt

image copied from the Primitive Skills web site

I was directed to this set of videos by Jeff Evarts.
Jeff is a past student here (Bloomery Iron Smelting), and has an all round interest in ancient metalworking and related technologies. He posted the ‘Making Steel’ video to the ‘Iron Smelters of the World’ Facebook group, asking for opinions. (1)

Initially I watched the first video ’Steel from Ore’ (seen below), as available on YouTube.
The first time through I made about a page of notes.
The following day I dug deeper, hoping for some better information. This got me to the ‘Primitive Skills’ web site. This gave a second video, ‘Build Furnace’ (seen below) and a longer written description of the iron smelt process as a blog posting. Eventually I discovered a yet a third related video and description, of the piston bellows used.

Backing up step.

Just who is ‘Primitive Skills’?
http://primitiveskills.net/

" Primitive Skills; is the technique that “antediluvian” used to sustain life in the natural environment. These techniques are meant to provide the basic necessities for human life including food, water and shelter. "

quote from the Primitive Skills web site

Unfortunately, the combination of web site, blog postings and individual videos on the YouTube channel does not include any references to Who or Where.  (1)
Where : A few things can be deduced. It is clearly a tropical location. Bamboo is seen and used as a fundamental raw material. It appears the presenter is Indonesian / Philippinno? So a guess would be someplace over to SE Asia ??
Who : The presenter does not talk - at all. The blog is in English, but there are hints from the writing that this is not the primary language. (3)
This really appears to be a single person operation. The filming has the look of ’set up and walk into the frame, edit later’. (4) The demonstrator also certainly appears to be undertaking all the work involved himself, which is re-enforced by descriptions in the blog posting. This to me suggests an individual with a fairly good level of overall tool use skills. There are technical descriptions on the blog post detailing at least some understanding of the bloomery iron making process. So at the least some background research was undertaken.

One important limitation is that the sequences are ‘live - as it happens’. This means camera angles are sometimes not the best, or things just did not get filmed.
I consider it fundamentally important here to also consider the overall intent of the series. It is not necessarily to demonstrate the ‘absolute best’ methods. It is clear that the combined series is intended to record a progression of nothing to something - ‘Walk into the bush and make everything you might need’. (5)

A) ‘Build furnace to prepare for iron age’



I was not able to find a supporting blog post for this sequence - which might have provided some of the details only estimated here

Although I did watch this second, I think looking at it first will help with better understanding.

There is a very clear ‘chicken or egg’ thing going on. Initially the work is gathering bamboo, using a simple triangular bladed *iron* axe. Which comes first? So move past this, and figure this just as easily (with considerable more work!) could have been one of various stone tools, perhaps even copper or bronze. The series presenter does describe this whole series as ‘moving from the stone age to the iron age’.

The first third of the video puzzled me at first (the lack of commentary a problem). The demonstrator is shown cutting bamboo, preparing splints. At first, given the title, I though this was related to making an interior framing for the furnace build. In actual fact this was illustrating ‘making tools to gather materials’, in this case one of several baskets that get employed later. Not really important to the furnace build specifically.

The build is shown using locally gathered clay, then mixing and combining with some amount of grass, as long strands of flat blades. The exact proportions here can not be determined, but it appears that there is not much of this organic material in the mix. This suggests the grass (especially given its flat profile) will be acting only as a re-enforcing element. I have to mention that this looks like very nice clay, as he is able to use it straight out of the ground with only some extra water added. There is no indication that additional sand is added (to heat stabilize or increase temperature resistance).
Overall the clay mix appears fairly wet during the build process.

The first step in the build is placing a flat disk of clay (looks about 3 cm thick) down as a hard base. Honestly, I have no idea why this was done. Considerable experience has proven that the underlaying base ground has little effect on the operation of the furnace. If anything, placing a layer of clay under the furnace is likely to create more problems for later slag tapping or extraction of the final bloom.  The only thing I can think of here is that this may be some (poor) reflection of the build method seen in Japanese Tatara furnaces? (6)

At about 8:30, you can see the first course of wall constructed :
- The overall interior diameter of the furnace base (not given) appears to be about 20 cm.
- The walls are very thick, looking about 10 cm + thick at the base. 
As the build progresses, you can tell the clay mix has changed to a firmer consistency. About half way through building up, a gentle wood fire is started, eventually to grow to the entire volume filled with sticks. (Likely to firm up the lower clay sections before completing the build). There is then a sharp cut in the action, with next the furnace shown at its full completed height and form. This appears to be twice the base diameter, so another WAG + is to about 80 cm tall.

The tuyeres appear to be hand built. They also are a significantly different colour than the raw clay (even later when that is seen dried). This certainly suggests these have been pre-fired into ceramic. There is no indication about this given.

image is screen capture from 'Build Furnace'

Two clay tuyeres are seen :
- about 20 cm long
- interior diameter looks about 2 - 2.5 cm
- very thick walls, about 2 cm
You can see as the tuyere is added that it is placed very low, with little clearance above the hard floor.
At one point (9:10) there is a top down view, which shows the distance proud of the interior wall is about 1/3 the distance across the furnace, so roughly 6.5 cm proud.

image is screen capture from 'Build Furnace'

Of special note is a detail of the interior of the finished furnace, taken later in this sequence (at about 12:40).

image is screen capture from 'Build Furnace'

You can see the placement of the tuyere, the image shot through what would normally be the tap arch : 
- The angle of the tuyere is extremely flat, near as I can measure off the screen at less than 10 degrees.
- The height of the tuyere is at best equal to the it’s width. That would place the centre mark at best at about 6 - 7 cm above the hard base.


B) ‘Piston Bellows (Fuigo)’


There is another separate video sequence and blog post illustrating the construction of the double action piston bellows used.
http://primitiveskills.net/primitive-skills-piston-bellows-fuigo/

The whole bellows is made from a section of hollowed out log.
These are the measurements quoted on the blog section
Total Length = 100 cm
Total Exterior Diameter = 30 cm
Total Interior Diameter ≠ 20 cm
Working Depth (distance seen as stroke during operation) ≠ 40 cm

So a real WAG here on air produced is about 12.5 litres per stroke .
Stroke speed appears to be about 1 per second, so double WAG at 750 LpM (??)


C) ‘Making Steel From Iron Ore’

With the related blog posting
http://primitiveskills.net/primitive-skills-making-steel-from-iron-ore/

Again, the limited nature of details in the written commentary, especially coupled with the lack of narration in the video, limits clear understanding.

Initially, my eyebrows raised at ’Steel’.
Grumble, Grumble = Bloomery IRON is not Steel.
Reading the related blog posts however, did clearly illustrate that the individual behind Primitive Skills :
a) was not a primary English language speaker (3)
b) actually had some understanding of the chemistry and physical process involved

So - Notes on what is seen (rough film times) :

0:00 - 5:00 = Five minutes of gathering rocks.
This certainly appears to be a limonite Fe203 ore. The blog posting gives “iron composition at 53% to 56%” Which, if actual Fe content, is a pretty good ore.
My first question here is - ‘Why such such big pieces?’
The ore is obviously surface deposited, eroded out and misplaced from the bedrock underneath. Big pieces have not moved as far down hill from source. Are there not smaller pieces further downhill or along the stream bed seen at other points? Sure, you have less pieces to pick up - but there is sure to be a penalty in the ore preparation phase.
In the blog there is a description of only being able to haul 35 - 50 kg ore per ‘turn’, with ‘several times to bring it back’. This appears the only indication of how much ore was at least gathered.

5:00 - 7:00 = Two minutes of gathering logs, cutting lengths, splitting logs.
This really puzzled me. At first I though this was the start of charcoal making. So, ‘Why not use the branches you had obviously trimmed off?’. The real purpose here comes a bit later.

7:00 - 8:00 = Breaking big rocks into ‘half hand’ sized pieces.
There does not appear to be any sorting going on here. Obviously a lot of hard labour here (use of a rock anvil and hammer stone).

8:00 - 10:00 = Roasting ore.
So now you see what the segment of chopping and splitting wood was all about.
There is a nice sequence at about 9:20 where one of the roasted pieces is seen easily broken smaller - just by hand.
But honestly? THAT is why you roast a rock based ore in the first place! Roast, THEN break. Certainly smaller pieces are going to heat quicker to the critical temperature (convert to Fe3O4), but still…

10:00 - 10:45 = Breaking ore.
The majority of the effective work here is seen using a hardwood length as a mallet over a stone block. The particle size appears what I would consider standard for a limonite ore, being ‘rice to pea plus the dust’. If anything, the particles may be a bit on the small size. There does not seem to be much in terms of rock / silica impurities included.
There is a sequence showing the use of the water powered ‘hammer mill’ built in a different sequence. Again I can’t see really how effective this really was. The stroke rate is very slow, with a very reduced impact force. The wooden hammer mill does not appear to be having much effect on the rock pieces under it as shown.
In the blog he mentions it taking ‘a day and a night’ under the hammer mill to prepare all the ore. Honestly, in my own experience, he would have been much quicker just to have stayed with the hand mallet.

10:45 - 12:15 = Equipment and Materials
This sequence is a pan over the various materials, equipment and tools prepared and on hand for the smelt itself.
The clay tuyere is seen buried in a heavy ring of clay around the base of the furnace. This helps link the hollow bamboo pipe up to the piston bellows. The clay appears to add about an additional 10 - 15 cm thick band, about 10 cm tall, around the base.

10:45 + 11:45 = Ore quantity ??
I mention these quick shots, mainly because there is no actual data given on any of the quantities used in the smelt.
It would be an extremely rough guess (at best) to attempt to estimate how much ore was prepared, or if in fact this was all the ore actually used. This appears to be something about enough to fill  3/4 to 1 standard 20 litre plastic pail? My best guess is that this would run to about 40 - 45 kg of the ore (??)  (7)

11:45 12:00 = Charcoal
You see the prepared charcoal. The particle size here looks about the same as I would suggest, roughly ‘egg / golf ball’ to about ‘pea’.
There is mention in the blog that this charcoal was ’stored from the previous year’. Given the high moisture environment seen in much of the video footage overall, I would wonder just how much additional water that fuel had soaked up over all that time. (Use of fresh and bone dry charcoal has proved to make a difference in our own work here.)

12:15 - 12:40 = Plugging the ‘tap arch’
I have to admit, this really puzzled me.
The furnace was initially constructed with a tap arch, although this is not shown as part of the initial construction of the base area of the furnace. It may have been cut in after the walls had firmed up?
The size looks roughly 10 cm tall by at best about 15 cm wide. The arch sits above that heavy ring of additional clay. This places the bottom of this *possible* tap arch certainly higher in the furnace than the tuyere (likely about even with the tuyere OD top edge).
We see the arch being completely sealed with a very heavy addition of clay, packed and blended into the thick base walls.

12:45 - 14:00 = Filling and igniting the stack
You can see a long sequence of getting the interior filled and fully ignited. A few burning coals are added, topped with that appears the ‘standard measure’ of charcoal. Then the single (!) worker takes a turn at the bellows to get that fuel fully ignited, and the process repeated. Although again very hard to estimate, the amount of charcoal looks to be about half one of our standard buckets, maybe about 1 kg per addition (? again there is no description provided).
As the last fill to top is added, you can clearly see the large amount of very white ‘smoke’. This likely is as much steam as anything else (remember the comment about damp fuel). You can see the change at the top of the stack (smoke disappears), as the furnace becomes ready for the first ore charges.

14:00 - 14:40 : Charging Ore
Again, without numbers at best there are only WAG’s. It looks like about 500 gms  (?) or so of ore being added. The level at the top of the furnace is seen dropped perhaps 10 cm at most. This looks to be a pretty standard addition proportion.  You can also see the full ignition to a clear flame a the top of the furnace.
There is very clear reference made on the written description of the importance of smaller charges  of ore and charcoal, to evenly distribute ore though the working stack in the furnace.

14:40 - 15:30 = Main Sequence
Just a note on the edit. I thought the large set of short clips showing ‘more of the same’, especially the distance shots of bellows pumping, really helped to enforce the length of the smelting sequence itself. No specific note is made in the blog of how long the primary sequence to burning down took.
A reference is made to “After a hard working day…”

15:15 = Steam Venting?
Something I noticed here, which re-enforced an aspect I wondered about, and had been hinted at during the ‘block the tap arch’ sequence.
There was no mention of a drying fire (visual or in the blog). The furnace walls certainly appeared quite damp. At this point you see an image taken that clearly shows steam venting off the side walls of the furnace. With such thick walls, the danger of failure due to cracking never became a problem. But the amount of energy being robbed by driving off all the moisture in that very massive clay construction must have been considerable. Normally I would remain concerned that this would end up effecting possible internal temperatures, so the overall progress of the smelt.

15:50 = Tamping Down
The top surface of charcoal is seen being tapped down inside the furnace, which appears to have burned down to about half full.
This may be an attempt to tell just where the top of the slag mass is. The sound seems more like crushing charcoal to me, but in truth this is as much a ‘feel’ thing as a change in sound.

15:30 - 15:45 = Clearing Base
16:00 - 1630 =  Open Tap Arch
In the first sequence we see that the heavy ring of clay around the base has been broken free and removed. The exterior of the furnace is being carefully fractured in preparation for extraction (?)
In the second, the fill to the tap arch is broken out, exposing the slag mass within. There is no actual liquid slag tapped. This may be because of the thick wooden tool with only a squat point being used. This is seen not being effective in breaking through the solid bottom of the slag bowl.

Another significant problem is clear here. As a single person operation, including use of a manual bellows, the worker can either tend the furnace, or keep air flow and as a result operating temperatures maintained. Clearly, the whole furnace interior is starting to freeze up. (8)

16:30 - 17:00 = Extraction
The walls of the furnace are broken away, exposing the entire slag mass within
Near the end of this sequence, you can see how the slag has ran down to, and then fused completely, to that hard clay base built into the furnace.
Again, the complete absence of any liquid slag indicates the temperature in the entire mass has dropped below ‘about’ 1100 C (9)

17:00 - 17:35 = Freeing the ‘Bloom’
Clearly there is one solid block of slag, still with fragments of clay wall clinging to it. The brittle nature of the slag is clear by the way it shatters under the blows from the large wooden  (troll) hammer.
Eventually the much smaller mass of metal and clinging slag is broken out - only a small fraction of that large mass.

17:35 - 18:55 = ‘Compaction’ ?
(see the comment on video and visual temperature)
That there is an iron bloom created is not in doubt.
The material is entirely too cold at this point for any real effective compaction. The fact that the material collapses over the first strokes of the troll hammer is more an indication of how lacy and slag filled the results are. Dark slag is certainly seen breaking off, and remaining crusted on to the exterior.

18:55 - 1920 = Cold Hammering
An attempt is made to compress the surface further using a hand stone on a rock block. The metal is certainly far too cold at this point, both image and sound clearly show this.
I expect this was included primarily to complete the working process for the video.

The final sequence that shows the stone polishing of the bloom has clearly been added to ‘show there was bright metal’ - so an actual iron was created. The final close up images do show a considerable amount of visibly fluid slag still remaining.

image is screen capture from 'Build Furnace'

It is extremely difficult to estimate much about the results of this effort.
The mass appears to be about 15 x 20 x 10 cm, but it is extremely lacy and still includes a lot of slag.
My best estimate is, once that mass has been correctly re-heated and compressed down to a working bar, there will be significant loss. I would expect the end result will likely be about enough solid iron to make the triangular inset axe blade seen used in the first part of this series.

The main correction I would suggest is with the initial layout of the base and tuyere placement of the furnace.
- There is no need for that solid clay base. If anything this compounded slag control problems.
- The tuyere was set far too low in the furnace base. There simply was no room for slag to collect.
- The tuyere angle was far too shallow
- Taken together, this meant the slag bowl was not deep enough, not allowing the metallic iron to collect most effectively into a more solid bloom.
- With a working tap arch, (and depth to work) it might have been possible to tap off the obviously excess about of slag. This has been shown to help create a more solid bloom. (10)

Likely this also has proven that undertaking an iron smelt, especially if intending human powered air is most certainly not a one person activity! (11)

Taken altogether, the video certainly does illustrate the use of a primitive process to work through from raw ore to iron production. I certainly realize, as all viewers should, this was never intended to be a tutorial on ‘best method’.

Honestly, ‘Primitive Skills’ should be applauded for actually creating an iron bloom - working not only using such simple methods, but with what looks to be a first attempt, and most certainly by undertaking such massive effort working alone!


Notes:

(1) I have recently been criticized as being 'needlessly pedantic'.
If you are going to consider this report on my own observations, placed against experience, as stating the obvious and far too detailed : Please just turn to some other commentary.
It remains my experience, over a number of projects and decades of teaching and experimentation, that is is always better to state the obvious, than assume others know the details. Too many times I have seen problems caused by 'I thought you knew that', assumed in error.
I generally assume that most readers are not in fact familiar with the large number of earlier descriptive commentaries available both on this blog and over on the main web site documentation.

(2) I placed a comment on to the related blog posting, both making some technical observations and also suggesting that some description of the individual involved and the working location would really be helpful in assessing the information provided.

(3) I most certainly suggest anyone following the links (and I recommend reading the fuller blog description of the smelt attempt) be very considerate of the writing style. I see a few European usages (like ‘coal’ when ‘charcoal’ is meant).

(4) This itself not intended as a criticism. I use the same film technique myself - I suspect for pretty much the same reason. There is also every indication throughout that this is a totally one person production.

(5) This concept, of starting from bare hands and no tools, through gathering and processing natural raw materials through to production of all elements needed to create a finished object, now has the term ’Sole Authorship’ applied to it. I was first introduced to this as a recognized, systematic, process by Mike McCarthy (Mike was the instigator for the original ‘Gangue aux Fer’ and the first Early Iron Symposium.)

(6) In a Tatara, the initial clay construction at the base is fully intended to melt away during the smelting process. The rectangular shaped Tatara is also designed to make best use of fine ‘iron sand’ ores found in Japan. The final iron produced (called tamahagane) is log shaped, rather than the ‘half sphere’ typical in a round ‘short shaft’ furnace.
https://en.wikipedia.org/wiki/Tatara_(furnace)

(7) We worked a lot with a similar (looking?) limonite ore (‘Lexington Brown’) in the early years of our experiments here. So this estimate is based on those smelts, where typically we used about the same volume of ore - in the ranges of 30 - 35 kg total.
Please remember that is only based on the images - there is no actual data provided!

(8) A very important thing to remember. The camera sees heat colours differently than your eye! Generally, this is an amount equal to ‘one temperature grade’, so when you see orange on film, the actual heat colour is into the reds. A video seen as red is actually dropped at best into the ‘critical’ (below forgeable) range.
This is seen most clearly during the sequence where compaction is being attempted. The iron here is obviously well below actual forging temperatures.

(9) Again intended as a very rough measurement (only!) The solidification temperature of any slag is greatly modified by its iron content. This aspect is often intentionally modified (addition of either more silica via sand or iron oxide via hammer scale) by the most experienced iron makers.
I freely admit that slag chemistry is certainly my worst area of understanding.

(10) Think about slag containing metal particles, running through a sponge (that initial lacy bloom mass). As the slag runs through, it will leave bits of metal behind, slowly filling all the gaps.
( No - this is not exactly how it works, but more take the concept.)

(11) I did do a complete smelt by myself - once. Key here is *once*. And I had an electric blower for air.  (see : Redemption Bloom )